An evaluation of the Personality Assessment Inventory for personnel selection : a study with the New Zealand police : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in Psychology at Massey University

Abstract

The Personality Assessment Inventory (PAI) was evaluated as an instrument for New Zealand police selection. Recruits (N = 127) attending training were administered the PAI, and the data used to assess reliability, validity, impact of response sets, test fairness, and utility for selection purposes. The PAI was found to be less internally consistent than in other studies, but most scales achieved acceptable alphas. The PAI showed evidence of construct validity when correlated with the NEO-PI-R. PAI scores accounted for 12% of the variance in final grade achieved during training, compared to 8% for the NEO-PI-R. Defensive responding was identified as a considerable threat to validity. Small, but significant differences were found between demographic groups on some scales based on age and ethnicity. While the PAI might be a useful addition to the New Zealand Police test battery, the presence of less than ideal reliability, possibly high levels of defensive responding, and differences between demographic groups means that hypotheses generated from test scores will need to be carefully integrated with other sources of information.